1. Field of the Invention
The present invention relates to the field of signal processing systems, and in particular to the field of waveform shaping in optical information recording systems. More specifically, the present invention relates to a waveform shaping system for improving resolution of the recorded indicia in an optical disc while overcoming duty cycle and asymmetry problems associated with prior art techniques.
2. Brief Descriotion of the Prior Art
A great many systems are known for optically recording information on light responsive media. Digital information as well as analog information can be recorded on a variety of media, including photoresists, photochromic materials, and thermally responsive materials in the form of discs, drums, and tape.
Virtually all of these known systems employ some form of pulse code modulation (PCM) or frequency modulation (FM) due to the nonlinearities of known optical recording media. Additionally, virtually all of these known systems employ an optical system including an objective lens for focusing the writing beam of light to a small spot on the light sensitive surface of the medium. In order to maximize the information density of the recorded material, it is necessary to focus the writing beam to the smallest possible spot size. Due to diffraction effects, a focussed beam of coherent radiation forms a central bright spot surrounded by concentric spaced rings of light known as an Airy disc, having an approximately Gaussian power distribution. The diameter of the central spot of light is defined by the wavelength of the recording light and the numerical aperture (NA) of the objective lens.
Most optical recording media exhibit a threshold effect, meaning that a level of writing beam power density exists above which the medium will be altered, and below which the medium will not be altered. Since the power density of the focussed writing beam varies as a function of the distance from the center of the central spot, it is possible to form altered areas on a medium which are smaller than the diameter of the central spot of the focussed light. In order to obtain an altered area of a specific size, the prior art has taught that it is merely necessary to adjust the peak intensity of the modulated beam such that the power density of the focussed spot at the specified diameter is equal to threshold level of the recording medium. Experimental results indicate that this technique works extremely well when there is no relative movement between the writing spot and the medium and no thermal conductivity effects diffusing the energy absorbed by the medium. In all practical systems, however, relative movement between the medium and the recording spot is necessary. This relative movement greatly complicates the calculation of the amount of energy absorbed by a particular region of the medium.
All light responsive media exhibit a tendency to integrate the amount of energy received over a period of time, so that the threshold level of the material does not define a specific instantaneous power density at which the material is altered, but rather defines an energy density which is the result of an integration of the received power over the time during which the power is received. In other words, the threshold level for the materials is actually an exposure level. Exposure level is defined as the radiant flux per unit area integrated over the exposure time. The radiant flux irradiating a particular point on the medium at a particular time is a function of the position of the point within the focussed spot, as well as the instantaneous intensity of the writing beam. Specifically, the exposure level at a particular point on the medium is equal to the convolution integral of the instantaneous power of the modulated writing beam as a function of time and the power density distribution function defined by the path and speed of the particular point of the medium through the focussed spot, evaluated over the exposure time. Since the power distribution of the focussed spot varies in two dimensions, the exposure level over the medium defines a three dimensional graph. The intersection points of this graph with a planar "slice" through the graph at the altitude corresponding to the threshold level of the medium maps the boundaries of the altered areas on the medium.
An important concept in determining the effect of the finite size of the recording spot is the concept of the spatial frequency of the information on the moving recording medium. This is the subject of U.S. Pat. No. 4,616,356 which is incorporated herein by reference.
Also incorporated herein by reference is U.S. Pat. No. 4,225,873 describing signal processing techniques to achieve proper duty cycle of the recorded information and to minimize second harmonic distortion of the recorded information which are related parameters and which are improved by the present invention. In this description, the duty cycle of the recorded information will be referenced in detail, it being understood that second harmonic distortion is proportional thereto. That is, when duty cycle is controlled to approximately 50%, this minimizes the second harmonic distortion of the recorded information.
As previously indicated, when recording on a medium using laser light, the effects of the recording process may either be an ablation of the recording surface, a discoloration of the surface, a photo-hardening or a photo-softening of a photoresist coated surface, a raised bump or bubble formation, and the like. In all cases, a recording medium is moving relative to an impinging light beam, and the energy of the light beam is increased and decreased about a threshold level so as to cause or not cause the desired effect, respectively. Also in all cases, when the light beam is first turned on or caused to exceed threshold, the full effect of the energy on the recording surface is not instantaneously realized. Rather, and especially with mediums that are thermally responsive, the recorded effect (i.e. indicia) increases with time due primarily to the temperature of the medium increasing with time under the influence of the applied energy. That is, when the light beam is first turned on, instantaneously no alteration in a moving recording surface can be detected. If left on continuously, the full effect of the impinging light beam will cause a wide stripe of the effect on the recording surface, thereby defining a broadening of the track of the light beam with the passage of time.
On the other hand, when the light beam is turned off or falls below threshold level, the effect of the lack of sufficient energy from the light beam is almost immediately recognized by the disc surface, since the downstream portion of the disc is cold and unaltered. Accordingly, when the light beam is turned off quickly, an almost instantaneous stoppage of the recording effect takes place, and since the light beam is substantially circular at the point of impingement on the recording surface, a rather blunt end on the trailing edge of the recorded indicia can be observed.
If the moving recording surface is therefore impinged by a light beam having a sharp rising leading edge and a sharp falling trailing edge, a rather "pear-shaped" or "teardrop-shaped" surface effect will be noted, the leading edge being tapered and widening to a constant width until the beam was shut off, at which point the recorded effect would define a rounded or blunted trailing end.
While the aforementioned characteristics of the recorded surface results regardless of the recording "effect" chosen, for ease of discussion in this description, it will be assumed that the recording light beam causes an ablation of the surface, thereby producing a "pit". It is to be understood, however, that a bump, discoloration, photoresist hardening, photoresist softening, or other indicia can be substituted for the term "pit" without departing from the nature of the technical description of the recording effect. It is further to be recognized that the term "pit" as used herein would be equivalent to the bump or pit formed after development of a photoresist if that were chosen for the recording surface. Finally, the term "recording surface" refers to that part of the disc which is sensitive to the impingement of light energy, whether it be on the outer surface of the disc or submerged beneath the surface being impinged by the light beam.
One of the most important parameters in the recording of a CD (Compact Disc) master is the precise control of the pit-to-land ratio. Each transition from land to pit and vice versa marks a zero crossing of the EFM (Eight-to-Fourteen) signal; therefore, the size of the land area holds as much digital information as the size of the corresponding pits. The ratio of pit to land size is the "duty cycle" of the recording. A duty cycle of 50% would represent a symmetrical signal with pits occupying 50% of the track area and land area the other 50%. Any recording that deviates from a 50% duty cycle would be represented in an eye-pattern (discussed in connection with FIG. 2) as an asymmetrical signal. Due to limitations in the replication process, differences in the various CD player's optics, and variances in the players electronic boost, all CD recordings, even ones that were recorded with perfect 50-50 symmetry, may be seen by the playback electronics as slightly asymmetrical. Circuitry within the CD player corrects for this asymmetry by shifting the "zero level" to a point where no timing differences between positive and negative half cycles exist on average.
The optical head of a player "reads" the information on a disc by focussing a laser beam to a spot about 1 micron in diameter and scanning this spot along a spiral track of pits and lands on the disc. The lands reflect most of the light back to a photodetector in the head, but the pits reflect very little light. The photodetector generates an electrical signal proportional to the amount of light reflected by the area of the disc illuminated by the spot. Because the spot is diffraction limited and is only slightly smaller in diameter than the length of the shortest length pit on the disc, the rise and fall times of the electrical signal are relatively rather slow, and the amplitude of the signal representing the smallest pits is smaller than the amplitude of the signals representing the longer pits. The difference in amplitude varies from player to player because of differences in optical quality, and from disc to disc because of pit geometry. All players partially make up the difference in amplitude with an analog aperture compensation circuit which boosts the amplitude of the high frequency signals relative to the low frequency signals. Unfortunately, in the interests of economy, these boost circuits are not phase linear, so the harmonics of the signals are delayed by varying amounts, and the shape of the waveform is altered in addition to its amplitude.
The EFM signal recorded on the disc is a run length limited digital code which is self clocking and contains its digital information in the timing between transitions. On the disc, such transitions are the boundaries between pits and the lands between pits. Therefore, the length of each pit is a quantum of data and the length of each land between pits is also a quantum of data.
The rules for the EFM code according to CD standards require that each and every transition occur at one of nine allowed times after the last transition. More specifically, the period between any two transitions must be nT, where T is a fixed period of time (about 231 nanoseconds) and n is an integer between 3 and 11. In order to decode the information, the player must decide how many T's pass between each transition. The rate of rotation of the disc under the playback head is not stable enough to simply time the transitions with a fixed clock. Instead, a channel clock running at an average of one cycle per T (4.3218 Mhz) is phase locked to the recovered signal. This PLL (phase locked loop) operates by comparing the timing of each transition in the signal with the closest cycle end of the clock. If the transition consistently occurs just before a cycle end of the clock, the clock slowly speeds up until the transitions occur exactly at a cycle end.
Unfortunately, the rise and fall times of the analog signal from the disc are several T long, so a specific voltage must be chosen to define the instant of the transition. This voltage is the player's "Decision Level". The player chooses this level with the help of another rule of the EFM code which states that, on the average (and ideally), the EFM signal will be "high" for exactly the same amount of time, on average, that it is low. Knowing this, if a player picks a decision point too high in voltage, the EFM signal will be low (lower than the decision point) more often than it is high (higher than the decision point). A servo loop in the player slowly adjusts the limit voltage to find and maintain the proper decision level voltage.
It is important to realize that the player cannot change its decision level on a cycle by cycle basis, but the period between the transitions can vary from 3T to 11T instantly. Any frequency dependent influences on the shape of the waveform may cause the ideal decision point for the high frequency segments of the signal to be at a different voltage than the ideal decision point for the low frequency segments of the signal. The player will choose a decision point which is an average of these voltages. As mentioned earlier, the boost circuitry of the player is not phase linear and therefore causes a "spread" in the ideal decision point voltages with frequency.
In order to read the signal from the disc, the player must keep the small spot of light centered over the moving spiral track of pits on the disc. Various techniques are used for detecting the position of the spot relative to the track, and tracking devices of some players (using single-beam or push-pull trackers) are greatly influenced by the average duty cycle of the pits along the track. That is, these players will only reliably follow along the center of the track if the track consists of a certain percentage of pit area greater than the non-pit area (a duty cycle greater than 50%). While recorded intelligence information is contained in both the pits and the lands, no tracking information is contained in the lands which are coplaner with the other unrecorded regions of the disc surface, i.e. the surface of the disc between tracks of pits. Accordingly, the greater the length of the pit relative to the land between pits, the better the tracking subsystem works. As indicated, with any duty cycle less than about 50%, there is insufficient tracking information to enable push-pull trackers to stay on track. One solution to this dilemma, then, would be to increase the pit-to-land duty cycle to above 50% and rely upon the player's ability to readjust its "decision level" to compensate for the increase in recorded duty cycle. However, in order for the player's PLL to maintain a channel clock based on the EFM signal, the increase in the duty cycle must be uniform over the entire signal, but this does not happen because there is not an equal proportional increase in pit length over the nine different pit sizes (3T to 11T). A 3T pit will obviously have a greater geometric increase, percentage wise, than will a 6T or an 11T pit. Unfortunately, therefore, departing from a 50% duty cycle just to accommodate single-beam trackers sacrifices symmetry, and if asymmetry is too great, the player's phase lock loop may not be able to lock to the recovered signal.
Another consideration which has heretofore not been adequately provided for in the recording process is the fact that single-beam players whose tracking function depends on the amount of light deflected back to the photo-sensor in the direction of the tracking error operates best with tapered (pointed) ends on the pits formed in the recording surface. The ideal pit would be an elongated groove with an apex on the bottom of the pit, similar to the impression that would be made by an upside down pyramid. The more the slope of the sides and the more tapered the ends, the better. Accordingly, any attempt to improve the pit definition tends to create a flat plateau region and eliminate the long narrow and tapered leading edges of the pits making tracking more difficult for such single-beam players. Moreover, pit length varies greatly pit-to-pit, i.e., from 3T to 11T, and a 3T width pit or land can be placed directly adjacent to an 11T pit or land. If the peak power level is increased to produce rounded (less tapered) higher definition 3T pits, the same peak power level applied when forming an 11T pit would produce a very wide or blunted trailing end portion. This difference in geometry between 3T and 11T pits creates differential symmetry where the 3T pits have one symmetry and the 11T pits have another symmetry configuration, and this causes problems with player decoders. Ideally, there should be the same symmetry on each pit size. Otherwise, there will be a "spread in pattern" in the eye-pattern (discussed in connection with FIG. 2).
It can therefore be appreciated that without the background knowledge discussed above, a simplistic master recorder design would simply maintain a 50% duty cycle in the recorded information. In the ideal world, this would seem to be an obvious thing to do, since maintaining a 50% duty cycle would appear to be an essential requirement for the recording process. Methods and apparatuses of the distant prior art have thus merely provided a rectangular waveform in which the portion representing an nT length pit is equal in length to that portion of the waveform representing an nT length land between pits.
With experience, it has become evident that many factors must be considered in order to produce the optimal track of pits on the disc. For example, since the width of a pit along the track varies with the amount of time the beam is allowed to impinge on the surface, pit size and geometry is a function of surface speed--an important factor for discs recorded in a constant angular velocity format. Other factors to be considered, already discussed, are duty cycle (&gt;50%), the shape of the side walls of the pits in elevation (important for single-beam trackers), symmetry, and pit resolution.
Pulse length can be increased to raise the duty cycle, but this increases asymmetry. Increasing the power level of the modulated light beam improves resolution at the leading edge and increases duty cycle but produces "pear-shaped" trailing edges and excessive asymmetry. Increasing peak power at the leading edge by using a stepped modulator driving pulse improves resolution of the leading edge but requires a reduction of pulse length to keep the duty cycle under control. Doing so gives unequal symmetry for the different pulse lengths and requires an adjustment for each surface speed for best results. Threshold can be lowered to increase duty cycle, but this produces poorer resolution in the leading edges of the pits and ill-defined pit widths exhibiting substantial difference in geometrical shapes between a 3T and 11T pulse length. Finally, any attempts to increase pit resolution inherently diminishes the ability of single-beam trackers to track.
It can thus be appreciated that, while many adjustments of the parameters affecting the timing and geometry of the pits are available, the interaction by making any adjustment which affects other parameters is a tremendous problem which reduces yield and requires constant attention by a recording operator.
There is therefore a need in the art for a method and apparatus for driving the light modulator with a pulse which permits independent control over geometry and asymmetry over wider ranges than heretofore possible. The method and apparatus permits adjustment of pulse length and power to optimize asymmetry and permits adjustment of the slope of the trailing edge of the pulse to keep the geometry correct. This rather independent control of these parameters are extremely important when considering that for any given process for manufacturing discs, taking into account the differences in chemicals, surface thicknesses on the disc, molding machines, and mastering machines, any particular system would require a different adjustment to obtain optimal pit geometry and asymmetry, and the prior art has failed to provide this flexibility by being unable to control certain parameters independently. The present invention therefore satisfies a long need in the art for a method and apparatus which not only can permit optimal adjustment of the modulator driving pulse for a particular system but can also be adjustable to account for differences system-to-system. A synergistic effect is realized in the ability to produce elongated ellipse-shaped pits, i.e. pits with tapered ends, and yet maintain reasonably high resolution and required duty cycle, giving the benefit of improved tracking as well as improved signal recovery.